[PaulFolbrecht]

Hey Bob,

I thought I'd start a new thread since the other was no longer on topic.  It's a good discussion.

To recap, you correctly pointed out that I was all wet in suggesting that an amp could exceed the amount of current given by P=IV *at maximum gain (voltage)*.  But that was the only mistake in my rationale - at max gain, the amp can't put out more than P/V amps, but only at max gain is there such a restriction - and when are we ever listening at max gain?

Much more important is an amp's current-delivery capabilties at less than max gain, and this is theoretically (and, in practice, often substantially) completely unrelated to its maximum RMS wattage rating.  This is how we get very powerful amps that are rated for a rather meager amount of constant power delivery.

I can now say without hesitation that the Art Audio Carissa is one such amp.  I expected it to sound wonderful in the mids and highs, but I was flat out not expecting it to be able to manhandle my Gallos' woofers as it does.  Note that the Gallos do not qualify as hi-eff at 88 dB/W, though, as has been pointed out, their impedance and phase angle curves are fairly flat and so they are a fairly easy load.  I AM still planning on bi-amping, relieving the Carissa of the need to handle the lowest octave or two of bass, but honestly this wouldn't be stricly necessary, and have been conducting my first listening with no 2nd amp.  I am talking SPLs of 85dB at 2M from the speakers - not thunderous by any means.  YES, they do clip when pushed HARD - harder than I care to listen.

As for distortion degrading sound near clipping... yes, of course that's true.  SET amps in particular, of course, frequently look downright awful on paper - but still sound good.  BUT, we don't have to go there because this particular SET amp was built for low distortion across the entire audio range - under 1% at full power, not the 8%+ you might see at some frequencies near full power from many SET amps.  (2nd-order HD of as much as 4% as been shown to still be inaudible, and I don't that, but I wanted a SET amp with impressive distortion specs nonetheless.)

As far as whether this amp is appropriate for these speakers or whether I'd be better off with more power, let's run some numbers, assuming we want 85dB at the listening seat:

88 dB/W + 3dB for having two speakers -6dB for distance gives us exactly 85dB/W - thus, we need 1W continuous power from the amp for 85dB.  If we allow a factor of 15 for transients, the Carissa is still within its *continuous* RMS delivery capability, <1% distortion across the board.  We're not even considering the biamping now - we can assume at the very least another 3dB effeciency with 500W of solid-state power on the woofers.

On paper, then, we've got plenty of power.  In the real world, the Carissa is the best amp I have yet heard with my Gallos, not because it's powerful and dynamic (well, partly), but mostly because of it's ultra-clear and transparent mids and highs.  It has trumped:

50W/ch AES Six Pacs (definitely the runner-up)
200W/ch SS Marantz MA700s
150W/ch Bel Canto S300 (do somethings quite wonderfully but just not enough emotion for me).
200W/ch NuForce 9.02s

And the 22W/ch (triode) Cayin A88 - #3 after the Pacs.

Conclusions?

1) There is definitely something to the "tube power > SS power" arguments.  It's due to current delivery.
2) *I* prefer the sound of tubes, in general, and that sound is defintely *different* than SS amps.  (No arguments there, I'm sure.)
3) Bi-amping, with tubes (even SET) on top with SS muscle on the bottom for bass, is definitely a scheme that offers lots of advantages.  If you like tube/SET sound.

Cheers,
Paul

[Bob Reynolds]

To recap, you correctly pointed out that I was all wet in suggesting that an amp could exceed the amount of current given by P=IV *at maximum gain (voltage)*.  But that was the only mistake in my rationale - at max gain, the amp can't put out more than P/V amps, but only at max gain is there such a restriction - and when are we ever listening at max gain?

Sorry, but I don't understand the concept of max gain. My understanding is that an amp has a fixed gain, not a min nor a max gain. If you are talking about an amp's input sensistivity -- the input voltage that will cause the amp to produce its max power for a given load -- then I don't see how that's relevant. Since the gain is fixed, any input voltage lower than the input sensitivity spec of the amp will produce an output voltage less than its max and for the same load the current being delivered will be less.

Much more important is an amp's current-delivery capabilties at less than max gain, and this is theoretically (and, in practice, often substantially) completely unrelated to its maximum RMS wattage rating.  This is how we get very powerful amps that are rated for a rather meager amount of constant power delivery.

You'll have to provide an example, because I don't see it.

Here is a quote from Art Audio's web site on the Carissa:

Typically, amplifiers using a single 845 output valve per channel achieve a 24-26wpc power rating. By applying less voltage instead but drawing more current, we have designed an amplifier with less rated power. Operating in pure class 'A' to full power rating, we achieve lower-than-customary distortion specs across the whole bandwidth. Equally audible, we achieve far superior dynamics. Additionally, the lower plate voltage prolongs tube life.

My interpretation of this is that the designer has designed the amp with a greater amount of dynamic headroom than typical 845 designs by not running the tube as "hot". This does not imply that the maximum current delivery of this design is any greater than any other 845 design. It can't be. The fundamental limitation of current delivery is the 845 tube.

Since he's not pushing the tube as hard as in other designs his distortion specs at max power will of course be better (his max power will also be much lower). I doubt that they will be substantially different from any other 845 design at say 10 watts or 20 watts.

this particular SET amp was built for low distortion across the entire audio range - under 1% at full power

That's because full power for this amp does not push the tube and is substantially less than typical 845 designs.

As far as whether this amp is appropriate for these speakers or whether I'd be better off with more power, let's run some numbers, assuming we want 85dB at the listening seat:

88 dB/W + 3dB for having two speakers -6dB for distance gives us exactly 85dB/W - thus, we need 1W continuous power from the amp for 85dB.  If we allow a factor of 15 for transients, the Carissa is still within its *continuous* RMS delivery capability, <1% distortion across the board.  We're not even considering the biamping now - we can assume at the very least another 3dB effeciency with 500W of solid-state power on the woofers.

Sorry, but I'm not following the +3dB for two speakers nor the -6dB for the distance. For a point source (your speakers), the sound decay follows an inverse square law. Please provide more detail here, thanks.

1) There is definitely something to the "tube power > SS power" arguments.  It's due to current delivery.

That's definitely not true. Given the same power rating into the same load, the current delivery will be exactly the same regardless of the amplifying device. That perception is probably more related to the fact that tubes clip gracefully.

2) *I* prefer the sound of tubes, in general, and that sound is defintely *different* than SS amps.  (No arguments there, I'm sure.)

Agreed. There is no argument against personal preference. When I replaced an NAD C320BEE integrated amp in my office system with a Decware tube amp due to not having any level control (the NAD amp had way too much power), I expected to hear quite a bit of difference. I heard essentially none and honestly, was disappointed. My guess is that as long as the load on the amp is moderate (say 1/2 of its max power so their distortion figures are comparable), I doubt that tube vs SS could be determined blindly.

3) Bi-amping, with tubes (even SET) on top with SS muscle on the bottom for bass, is definitely a scheme that offers lots of advantages.  If you like tube/SET sound.

Gee, if tubes have so much more current available that SS, then you'd really want a tube amp on the bass, right? Seriously, I'm a fan of bi-amping and as long as the gain is matched (by electrical measurement not by ear), you should reap some benefit.

[PaulFolbrecht]

Hi Bob,

Sorry, but I don't understand the concept of max gain. My understanding is that an amp has a fixed gain, not a min nor a max gain. If you are talking about an amp's input sensistivity -- the input voltage that will cause the amp to produce its max power for a given load -- then I don't see how that's relevant. Since the gain is fixed, any input voltage lower than the input sensitivity spec of the amp will produce an output voltage less than its max and for the same load the current being delivered will be less.

Once again, I failed to state things properly.  You're right that gain doesn't change - I should have said "max voltage output", not "max gain".  Most of the time, of course, the amp's input signal is nowhere near it's full-power-sensitivity rating (for the Carissa, .611V), so the output voltage is also not near maximum.

AT THOSE TIMES, which, again, is almost all of the time, the amount of momentary current the amp can supply might vary greatly from another of similar RMS power rating.  I believe this is mostly a function of power supply, capacitance reserves, and, in the case of tube amps, the output tranformers.

Here is a quote from Art Audio's web site on the Carissa:
Typically, amplifiers using a single 845 output valve per channel achieve a 24-26wpc power rating. By applying less voltage instead but drawing more current, we have designed an amplifier with less rated power. Operating in pure class 'A' to full power rating, we achieve lower-than-customary distortion specs across the whole bandwidth. Equally audible, we achieve far superior dynamics. Additionally, the lower plate voltage prolongs tube life.

My interpretation of this is that the designer has designed the amp with a greater amount of dynamic headroom than typical 845 designs by not running the tube as "hot". This does not imply that the maximum current delivery of this design is any greater than any other 845 design. It can't be. The fundamental limitation of current delivery is the 845 tube.

Since he's not pushing the tube as hard as in other designs his distortion specs at max power will of course be better (his max power will also be much lower). I doubt that they will be substantially different from any other 845 design at say 10 watts or 20 watts.

1st, the amp does have better distortion figures across the board than most single-ended 845 designs.  There are other factors at work there.

As to the rest of what you said, I think that's accurate (that the 845 has an inherent max current capability).  I don't see how that's terribly relevant, tho.  You're certainly right that their able to able to provide more current because they lowered the voltage, but I don't believe that's the whole story. 

Sorry, but I'm not following the +3dB for two speakers nor the -6dB for the distance. For a point source (your speakers), the sound decay follows an inverse square law. Please provide more detail here, thanks.

Well, +3dB results from two speakers providing double the energy, no?!  You're right about the inverse-square law; I approximated 6dB because I didn't feel like doing the math.  I'm about 3M from the speakers so, you're right, a bit more may be lost.  There's also room-reinforcement effects, which are of course positive and I didn't consider.

Agreed. There is no argument against personal preference. When I replaced an NAD C320BEE integrated amp in my office system with a Decware tube amp due to not having any level control (the NAD amp had way too much power), I expected to hear quite a bit of difference. I heard essentially none and honestly, was disappointed. My guess is that as long as the load on the amp is moderate (say 1/2 of its max power so their distortion figures are comparable), I doubt that tube vs SS could be determined blindly.

Hmm.  I agree that some of the best SS amps can sound a lot like tubes.  From what I hear, the new FirstWatt F3 sounds exactly like a good SE tube amp.  However, in the more usual class AB SS vs. PP tubes, I personally can hear a different within a few seconds, every time.  I've done a lot of A-Bing. 

A few months ago I swapped the Onix valve amp I had in my 2nd system for a friend's Arcam a65 (I think) so that he could hear tubes.  The differences were STRIKING, I must say, and, furthermore, at 38W and 40W rated RMS, respectively, there was absolutely no contest as to which had more drive.  This always seems to be the case... I'm not sure why... part of it is probably the very obnoxious clipping behavior of SS, but even when not pushed to clipped the equivilent wattage tube amps sounds more powerful than the SS amp.  Maybe the Onix just sounded so much better because it was just a better amp, but the Arcam sounded just like all the other non-hugely-expensively SS amps I've heard: impoper attacks and decays (too quick), exaggerated sibilants, collapsed soundstage.

3) Bi-amping, with tubes (even SET) on top with SS muscle on the bottom for bass, is definitely a scheme that offers lots of advantages.  If you like tube/SET sound.
Gee, if tubes have so much more current available that SS, then you'd really want a tube amp on the bass, right? Seriously, I'm a fan of bi-amping and as long as the gain is matched (by electrical measurement not by ear), you should reap some benefit.

That's what I thought as I was typing it, and realized it didn't make sense.  Damping factor is a big issue here too - tubes usually don't have it.  Maybe that's most of the story of why SS is generally stronger in the bass.  I swear, though, a 2nd Carissa on the bass here would do wonders!!  But that would be a very foolish move financially.

I don't think we're really "arguing" about much here, Bob, though it does surprise me that you don't hear much different in SS vs. tubes.  I stated incorrectly what I was initially trying to get across, and I thank you for setting me straight on that.

Going back to amping the Gallos - what started this discussion - you pointed out that the manufacturer recommends >=50W.  Well, that's typical low-cost SS power they're talking about, and in terms of real-world volume, and drive, no, it AIN'T the same as a good tube amp.  There's no SS amp in the world rated at 15W, or even 40W, that would sound as good (BE as good), as this tube amp I'm using now.  It's not the peak RMS number that's so important, but the instanteneous current delivery at lower voltages.  Max wattage is about as useful a metric of amplifier power as total harmonic distortion is of sound quality.  That's the way it is - in the real world, amplifiers are complex devices with a lot of other, usually unmentioned, performance characteristics.

I was happy to see, reading the Gallo SA sub amp manual, that they specifically discuss (and recommend, if you like it) using a "low-powered SET amp" on the Gallos as long as you're bi-amping to take most of the bass load off that amp.  (And, I still don't know EXACTLY how the Gallo scheme breaks up the load, because the top pegs do still drive the woofers, but if you've got a bass amp on the 2nd coils and especially if you turn up the crossover a bit, most of the load is assumed by the bass amp.)

If anybode else would like to chime in here, go for it!

Paul (Now happily set up amplifier-wise for my Gallos.. for more than a few months.. I hope.)

[Bob Reynolds]

1st, the amp does have better distortion figures across the board than most single-ended 845 designs.  There are other factors at work there.

As to the rest of what you said, I think that's accurate (that the 845 has an inherent max current capability).  I don't see how that's terribly relevant, tho.  You're certainly right that their able to able to provide more current because they lowered the voltage, but I don't believe that's the whole story. 

I did not mean anything negative in my comment about the Carissa's distortion specs. In my view he made a good decision to operate the tube well within its limits. He then has to fight a marketing battle against other designs that produce more power. He seems to have found a way by claiming increased current delivery. Though I still contend that he is limited by the chosen tube and in theory should not be any greater than any other design using the same tube.

Hmm.  I agree that some of the best SS amps can sound a lot like tubes.  From what I hear, the new FirstWatt F3 sounds exactly like a good SE tube amp.  However, in the more usual class AB SS vs. PP tubes, I personally can hear a different within a few seconds, every time.  I've done a lot of A-Bing. 

My own limited experience is that the better designed SS amps don't sound like tubes at all (thankfully) and that the better designed tube amps don't sound like SS at all (thankfully). These amps have very little sound of their own. That's as it should be; the amp (or any component) should not color the sound if truth is what you're after. That's not to say that there aren't tube amps that are catering to an audience that wants the traditional tube sound.

I don't think we're really "arguing" about much here, Bob, though it does surprise me that you don't hear much different in SS vs. tubes. 

I have heard differences, just the most recent exercise showed almost none. In changing from an entry Rotel integrated to a Jolida hybrid integrated I heard a big difference. The Jolida was very musical. I enjoyed it. Later I switched to a Krell integrated hoping for more detail and transparency. I got the detail and transparency, but the music died. Then moving to Bryston monoblocks gave me both -- the music returned without any loss of detail.

There's no SS amp in the world rated at 15W, or even 40W, that would sound as good (BE as good), as this tube amp I'm using now. 

That's a bold statement! You have done a *LOT* of A-Bing.

It's not the peak RMS number that's so important, but the instanteneous current delivery at lower voltages. 

It seems you're still confusing the lower voltage the tube is operating at with the voltage produced at the amp's output.

Max wattage is about as useful a metric of amplifier power as total harmonic distortion is of sound quality.  That's the way it is - in the real world, amplifiers are complex devices with a lot of other, usually unmentioned, performance characteristics.

Actually, max wattage into a given load, is exactly the metric of amplifier power.

I was happy to see, reading the Gallo SA sub amp manual, that they specifically discuss (and recommend, if you like it) using a "low-powered SET amp" on the Gallos as long as you're bi-amping to take most of the bass load off that amp.  (And, I still don't know EXACTLY how the Gallo scheme breaks up the load, because the top pegs do still drive the woofers, but if you've got a bass amp on the 2nd coils and especially if you turn up the crossover a bit, most of the load is assumed by the bass amp.)

Agreed. I would have preferred a single voice coil on the woofer and a clean separation.

Paul, I've enjoyed the discussion -- good to get the neurons firing every now and then. What are you doing preamp wise with the new amp?

Kind regards,

Bob

[PaulFolbrecht]

Hi again, Bob.  Sorry about my delay in responding - been busy.

As for what I meant by "tube sound" - truth is indeed what I was talking about, not any sort of coloration, "euphonic distortion", or any of that nonsense.  I didn't mean to imply "the special, different sound of tubes" by that phrase.  What I meant was that, in my experience, tubes usually do a better job of reproducing reality.  Of course, that's a very general statement, and as I think I said before I've never heard a great many of the expensive class A SS amps.

Quote from you: "It seems you're still confusing the lower voltage the tube is operating at with the voltage produced at the amp's output."

No, and I don't see where you got that from either.  I'm quite clear now that the amp's 25.5dB of gain means an increase in the input signal voltage of precisely that amount, at all times.

And

"Actually, max wattage into a given load, is exactly the metric of amplifier power."

Don't mean to be prissy, but you continue to use the textbook physics definition of "power" while apparantly being unwilling to simply recognize the somewhat different sense in which I'm using the term. :-}  I'm not arguing with the laws of phsics, Bob, but merely pointing out that an amplifier's power rating (max RMS wattage) is frequently NOT a good indication of its ability to drive loudspeakers with dynamics and authority and clarity to typical volume levels.  There is a lot more to the way an amplifier actually operates with a reactive load attached to it and a signal with great dynamic range than it's RMS power rating.  If this isn't getting across, I'd suggest reading the 6Moons reviews of the Carissa, the Hyperion 88 tube monos, a number of Positive Feedback reviews, and so on.  I'm using the same terminology and having the same experiences as these reviewers.  If you don't agree with their take on things or the way they describe them, fine - but we're not thumbing our noses at natural law or anything of the sort.  An amp can have a large wattage rating but a wimpy power supply that recovers poorly from transients or a number of other factors that can make it sound - and act - with less authority (current) than another with a comparitively smaller continuous RMS power rating. 

As I pointed out with numbers, with most of todays reasonably-efficient speakers, especially ones with very benign impedance and phase angles such as the Gallos, a couple watts of continuous power is all that's needed at typical levels.  Much more for transients, yes.

The linestage I'm now using is the excellent Sonic Euphoria PLC.  The extra gain from an active unit is not needed.  I'm typically around 1/2-2/3rds on the VC, so substantial attenuation of the signal is necessary for comfortable volume levels.

Cheers,
Paul

[PaulFolbrecht]

Bob,

Thought I'd let you know I was pushing the Carissa tonight and found that it does go into clipping at levels not too far beyond my usual listening levels. 

Every once in awhile, I do like to push it to 95dB+.  Based on how wonderful it sounds at 10dB lower, I wasn't expecting it to fall apart as quickly as it does, but it does.

So, um, hey, you were right all along.

Maybe.  I am not sure what to do now... frankly, I am TIRED of auditioning amps for the Gallos.  The Gallos must stay because I love them. 

At least I bought it used, at fair market value, and can get my money out of it (well, eventually) if I so desire.

[lonewolfny42]

How about a nice Butler 2250 for the Gallo's speakers....plenty of power. Found this on Audiogon..... http://cls.audiogon.com/cgi-bin/cls.pl?ampstube&1155784900  ...very good price !!

Butler info... http://www.butleraudio.com/tdb2250.php

[Bob Reynolds]

Thought I'd let you know I was pushing the Carissa tonight and found that it does go into clipping at levels not too far beyond my usual listening levels. 

Every once in awhile, I do like to push it to 95dB+.  Based on how wonderful it sounds at 10dB lower, I wasn't expecting it to fall apart as quickly as it does, but it does.

Hi Paul. I believe I read somewhere that 10dB is what most humans perceive to be twice as loud, so I don't think I'd be too disappointed about it. Every amp has its limits and it doesn't really look like it fell apart all that quickly to me. Remember that 3dB == double the power required, so you made a pretty hefty demand on the amp.

Maybe.  I am not sure what to do now... frankly, I am TIRED of auditioning amps for the Gallos.  The Gallos must stay because I love them. 

You are currently using inline high-pass filters in front of the Carissa, right? There's one option, but I don't know how game you'd be to trying it. It has some theoretical advantages. You might consider using an external active bass management controller with a good subwoofer. The active BMC has zero insertion loss unlike the passive inline filters you're currently using. Putting the bass source in a corner has some very real acoustic advantages, besides taking the load off the Carissa. The only downside that I can think of is that you pretty much reduce your Gallos to monitors, which actually may not be a disadvantage sonically.

Just for grins, read what Ken Kreisel has to say about bass management
http://www.mkprofessional.com/bass_mgmt.htm

I'm using this approach in both my home system and my office system. The office system uses a Decware 2wpc tube amp with 4 ohm nearfield monitors. It plays much louder than I can tolerate.

Best regards,
Bob

[PaulFolbrecht]

Actually, I have already taken the bass load off the Carissa using the 70hz highpass filters.  The Gallo sub amp handles the bass on its own.  Sounds very, very good this way - the Gallos do spectacular low bass, as tight and clean as you could imagine, down to 22hz.

Somewhat oddly, using the filters seems have very little effect on the loudness at which the Carissa clips.  Also amazing, the Carissa all on its own (no bass amp) does wonderful bass as well up to its limits.

I'm back looking at switchig amps again.  I didn't like the NuForces, but many ppl seem to like the CI stuff much better.

The utter simplicity of SET appeals to me intellectually very strongly but perhaps that is just something I need to get past. :-}

Paul

[Bob Reynolds]

Actually, I have already taken the bass load off the Carissa using the 70hz highpass filters.  The Gallo sub amp handles the bass on its own.  Sounds very, very good this way - the Gallos do spectacular low bass, as tight and clean as you could imagine, down to 22hz.

I was wrong to suggest the active crossover, it would be better than the passive filters, but neither can work well with the Reference 3. It's xover to the 1st coil of the woofer is at 150Hz, so having 70Hz or 80Hz high-pass filters in front of the Carissa will not completely take the woofer out of its path and that will always be a problem. This is just a limitation of a 3-way speaker with a single set of connections.

Since the speaker is designed to take a full range signal on its inputs, without modifying the speaker, it looks like more power is the answer.

Somewhat oddly, using the filters seems have very little effect on the loudness at which the Carissa clips.  Also amazing, the Carissa all on its own (no bass amp) does wonderful bass as well up to its limits.

That's because the woofer is still in play. Just curious, whose inline filters are you using? What's the slope?

I'm back looking at switchig amps again.  I didn't like the NuForces, but many ppl seem to like the CI stuff much better.

The utter simplicity of SET appeals to me intellectually very strongly but perhaps that is just something I need to get past. :-}

Just my opinion, but with your feelings towards tubes I wouldn't consider switching amps except Jeff Rowlands or maybe the new Bel Canto (they now use the same technology as Rowland). However, I would seriously consider higher power tube amps. Have you looked at the Rogue monoblocks?
http://www.rogueaudio.com/Products_m150.htm

Rogue offers good value compared to Audio Research and VTL.

-- Bob

[PaulFolbrecht]

Hi again Bob,

The filters are "FMod"s from HLabs.  I think 6dB/octave.  You may well be right about them not really hitting 22hz substantially with no room reinforcement - I've no idea.  Just know they definitely go lower with an amp on the bass inputs, no matter what you've got on top.

I have found that the 35W/ch Dared 845 SET amp I also have (for sale) can play MUCH louder w/out strain - in fact I got to 100dB, which frankly surprised the heck out of me.  So, it seems that I don't necessarily have to rule out 845 SETs.  The Consonance Cyber 845s I have thought about before... maybe.  I may also try the Ch. Island D200s.  I've tried the other new switching amps, which disappointed me though they are clear, dynamic, and detailed as heck, but not these.

If I was going to go PP tubes again, I'd go right back to the AES Six Pacs.  I think 0 neg feedback does wonders for any amp, PP, SET, or whatever.  When I first tried SET, I definitely discovered something I had not heard from any transistor amp or PP amp, and I've heard it now from more than one SET as well.  OTOH, I was never "disappointed" in the Six Pacs in any way and in fact loved them to death.  (In other words, if I wasn't a nut I'd have lived happily ever after and been done with it.  But, I knew I had to try SET and it gave me an experience I hadn't had before.)

So, like I said, maybe I'll try another 845 amp in the 30W range, and/or the CI blocks. 

The Dared can play plenty loud, though not sounding *quite* as good at the higher levels.  But, not pushed into clipping as was the Carissa.  Again, I don't like to listen that loud 80% of the time, but I need the extra power for that other 20%...

Later,
Paul

[PaulFolbrecht]

Oh, as another note, the Dared is rated at 35W/ch, although there's no doubt it's hitting double-digit distortion before it reaches that level.

AND, as another idea, I've been told (by N. Pass himself) that First Watt is doing a pair of 40W monoblocks.  Due the end of the year.  THOSE could be some killer amps!

[Bob Reynolds]

Maybe someday you'll get to hear a well designed high power amp (tube or SS).

Best of luck, Bob.