Caruso parametric curve

[trailerparkjesus]

What’s the best way to further correct the carusos frequency response? It’s close but needs some correction. Is it best to run white noise and auto correct with a eq matching plugin both channels at the same time? Maybe one channel at a time? Maybe not using a matching eq but just manually add more parametric bells in broad strokes against tones? Let me know what you do!

[Dub Studio]

When testing real-world response, I tend to record an impulse (one sample) and measure the response with EQ matching... that way you don't get all the surface noise and TT rumble skewing the result.

[markrob]

Hi,

I would think a single impulse measurement would have a very low S/N ratio. If you average many impulses you can improve the S/N , but that's not so easy to do. I guess if you window out the single impulse to be very short in duration, you would effectively filter out all of the low frequency noise, but you would also not have the full response. That might work if all you need it the high frequency response. That's probably true of the Caruso. A swept sine or MLS have really good S/N but you need some special processing to extract the impulse response. I've attached a couple a papers with some info. Each method had pros and cons.

AES134-000104.pdf

mls.pdf

I've been experimenting with a test signal with all of the 31 bands of the ISO 1/3 octave frequencies for head EQ. If you cut this, capture the playback and view with an FFT based RTA with decent resolution (64K points or so), each peak will align with a standard 31 band graphic EQ. You can then use the EQ to easily trim the playback for flat response. The problem here is that any harmonic distortion might align with other bands and cause errors in the response. Attached is a wav file with the test signal if you want to try experimenting with this. You should only cut as hot as you need to get above the noise floor to minimize any harmonic distortion bleeding. As long as you can see see the peaks, by a few dB, you are ok. If you need to cut hotter to get the highs out of the noise, you can brick wall high pass filter this signal to remove energy at the low frequencies. If you are cutting with RIAA, be careful here as the boost is quite excessive (cut at very low levels). Best to cut with no RIAA and either add it after playback capture or capture with a flat preamp.

3rd Oct Sin 20-20khz.zip

I prefer using regular program material vs white or pink noise when using something like Fab Filter because its hard to tell if the matched response is due to the noise floor. At least with program material, you can A/B with the source to make sure the EQ really matches.

Mark

[farmersplow]

Hi,
...
... Best to cut with no RIAA and either add it after playback capture or capture with a flat preamp.

3rd Oct Sin 20-20khz.zip

Mark

Cool test signal! I'll have to test that right away! Thank you Mark.

[flozki]

I would go very slow.cut some stuff you know and where you have the master and a reference on vinyl. listen on a decent system you know.compare and and do slight adjustments..
the fab filter .pink noise is the wrong way to go.
Technically yes .but soundwise very bad idea…
You don’t have to dream.
Caruso has a 25k price Tage difference to an sx74 when new.nowadays you need still 8-12k for a step more. So don’t go crazy.
What amp? QSC is a bad idea.many swapped from QSC to some better and quality increased dramatically.
How much fb? 28-29db makes it more linear but sound is much better around 26-27db this head has a sound and it sounds super nice if you know what you do
If you want flat professional mastering with no big work
You also need pro material.
Please don’t forget the purpose of Caruso head.it was and is still the only affordable alternative to overpriced and outdated Grampian heads.a nice no brained tune up for shitty VR heads and it beats most of existing westrex heads… so take it for what it is and dont push it with fab filter to something its not capable. You will destroy the head in short time.
Keep it simple.analog or at least on small changes in the 7-12k range and a few dBs change not more .
Happy cutting

[organicmusic]

I totally agree with Flo,
the Caruso head is a good sounding head for what it was made for. Buy a good amp and use your ears for measuring (it's still the best system to use, you can also add your stomach if in doubt...)

[boryo]

I totally agree with Flo,
the Caruso head is a good sounding head for what it was made for. Buy a good amp and use your ears for measuring (it's still the best system to use, you can also add your stomach if in doubt...)

please recommend a good model amp. i have qsc but obviously i will need better one. thanks!

[Dub Studio]

I would go very slow.cut some stuff you know and where you have the master and a reference on vinyl. listen on a decent system you know.compare and and do slight adjustments..
the fab filter .pink noise is the wrong way to go.
Technically yes .but soundwise very bad idea…
You don’t have to dream.
Caruso has a 25k price Tage difference to an sx74 when new.nowadays you need still 8-12k for a step more. So don’t go crazy.
What amp? QSC is a bad idea.many swapped from QSC to some better and quality increased dramatically.
How much fb? 28-29db makes it more linear but sound is much better around 26-27db this head has a sound and it sounds super nice if you know what you do
If you want flat professional mastering with no big work
You also need pro material.
Please don’t forget the purpose of Caruso head.it was and is still the only affordable alternative to overpriced and outdated Grampian heads.a nice no brained tune up for shitty VR heads and it beats most of existing westrex heads… so take it for what it is and dont push it with fab filter to something its not capable. You will destroy the head in short time.
Keep it simple.analog or at least on small changes in the 7-12k range and a few dBs change not more .
Happy cutting

Ok but let’s be honest, listening and making adjustments by ear is mastering, and that’s largely a subjective process.

If the OP wants to “correct” the head, we are not talking about mastering exactly, we are talking about a procedure we can use to evaluate the performance of the head before we start mastering… preferably a procedure that is measurable in some way so it can be replicated and / or adjusted.

Pink noise might not be the perfect way to do that, but it can certainly help to build up a picture, (and it has much less high frequency content than white noise, so it puts less strain on the cutting head) just remember to adjust your spectrum analyzer so that pink=flat.

I agree that extreme correction is probably going to cause more problems than it solves, so in that sense yes you will have to work within the limitations of the Caruso head. If your pink noise is telling you to make excessive boosts or cuts, you might have to concede that the head just isn’t capable of hitting your target.

Get it as close as you can with modest changes at first, and see how it performs in the real world. And yes, if it sounds a bit dull for example, you can always tweak it, but there is no harm in using some form of analysis to verify the results. You can learn a lot by observing and quantifying the changes you make... you don't always get out what you put in.

I also think its important to clarify that better (or more expensive) equipment does not mean a system will give better results, it just means it may have the potential to do so. I think its misleading to suggest that somewhere over the expense horizon there exists a kind of cutting shangris-la where you can achieve great cuts almost without listening. You always need to verify what you are cutting and you cannot buy better results, you have to make them happen.

I know the OP is asking about the Caruso head, but I think this is relevant to any head really. We are all part of a tiny community of people trying to make the most of the equipment we have (good or bad). So instead of drawing pointless distinctions between different classes of cutting head, perhaps we should think about some general advice that could apply to everyone.

[Dub Studio]

I would think a single impulse measurement would have a very low S/N ratio. If you average many impulses you can improve the S/N , but that's not so easy to do. I guess if you window out the single impulse to be very short in duration, you would effectively filter out all of the low frequency noise, but you would also not have the full response.

An impulse is quicker and cleaner than using a sine wave and contains all the low end info you need. It gives at least as good, if not better results.

I cut several pulses, and then record them back into the DAW. I then look at each one to see which is the cleanest, and then use that for the EQ matching. First I "top and tail" it to further reduce the surface noise interference, and then play it several times through the EQ matching software to capture the response. I then use that to record white noise / pink noise / more impulses / actual audio to verify the result. RInse and repeat until I am happy(est).

If you wanted to, I guess you could use several different pulses in rapid succession to average out the noise, it depends on how the software collects the data. I don't tend to bother because I don't find it makes any meaningful difference.

[Dub Studio]

What’s the best way to further correct the carusos frequency response? It’s close but needs some correction. Is it best to run white noise and auto correct with a eq matching plugin both channels at the same time? Maybe one channel at a time? Maybe not using a matching eq but just manually add more parametric bells in broad strokes against tones? Let me know what you do!

BTW I would avoid doing different corrections on each channel.

[misjah]

I would go very slow.cut some stuff you know and where you have the master and a reference on vinyl. listen on a decent system you know.compare and and do slight adjustments..
the fab filter .pink noise is the wrong way to go.
Technically yes .but soundwise very bad idea…
You don’t have to dream.
Caruso has a 25k price Tage difference to an sx74 when new.nowadays you need still 8-12k for a step more. So don’t go crazy.
What amp? QSC is a bad idea.many swapped from QSC to some better and quality increased dramatically.
How much fb? 28-29db makes it more linear but sound is much better around 26-27db this head has a sound and it sounds super nice if you know what you do
If you want flat professional mastering with no big work
You also need pro material.
Please don’t forget the purpose of Caruso head.it was and is still the only affordable alternative to overpriced and outdated Grampian heads.a nice no brained tune up for shitty VR heads and it beats most of existing westrex heads… so take it for what it is and dont push it with fab filter to something its not capable. You will destroy the head in short time.
Keep it simple.analog or at least on small changes in the 7-12k range and a few dBs change not more .
Happy cutting

Ok but let’s be honest, listening and making adjustments by ear is mastering, and that’s largely a subjective process.

If the OP wants to “correct” the head, we are not talking about mastering exactly, we are talking about a procedure we can use to evaluate the performance of the head before we start mastering… preferably a procedure that is measurable in some way so it can be replicated and / or adjusted.

Pink noise might not be the perfect way to do that, but it can certainly help to build up a picture, (and it has much less high frequency content than white noise, so it puts less strain on the cutting head) just remember to adjust your spectrum analyzer so that pink=flat.

I agree that extreme correction is probably going to cause more problems than it solves, so in that sense yes you will have to work within the limitations of the Caruso head. If your pink noise is telling you to make excessive boosts or cuts, you might have to concede that the head just isn’t capable of hitting your target.

Get it as close as you can with modest changes at first, and see how it performs in the real world. And yes, if it sounds a bit dull for example, you can always tweak it, but there is no harm in using some form of analysis to verify the results. You can learn a lot by observing and quantifying the changes you make... you don't always get out what you put in.

I also think its important to clarify that better (or more expensive) equipment does not mean a system will give better results, it just means it may have the potential to do so. I think its misleading to suggest that somewhere over the expense horizon there exists a kind of cutting shangris-la where you can achieve great cuts almost without listening. You always need to verify what you are cutting and you cannot buy better results, you have to make them happen.

I know the OP is asking about the Caruso head, but I think this is relevant to any head really. We are all part of a tiny community of people trying to make the most of the equipment we have (good or bad). So instead of drawing pointless distinctions between different classes of cutting head, perhaps we should think about some general advice that could apply to everyone.

i would listen to the guy who designed and make/made the Caruso heads

[flozki]

Pink noise might not be the perfect way to do that, but it can certainly help to build up a picture, (and it has much less high frequency content than white noise, so it puts less strain on the cutting head) just remember to adjust your spectrum analyzer so that pink=flat.

this is all fine. i totally agree to measure with whatever you have to understand your system.even with pro heads and high priced equipment. always.

but from the experience of broken heads returning for repair i just can tell that a lot of the users think that pressing the "correction" button on fab filter is the way to do and and its sorted out. that with a 500W per channel to push out the correction...
this destroys the head almost instantly. sometimes they think it works even better with RIAA curve on and +24dB output level of the DAC. AUTSCH!.
well shit happens if you dont know what you are doing. even with professional heads.

if you want NO measurable difference from your digital file to a vinyl record, better sell DSD streams or go back to the good old CD.
then measurement differences are at a minimum.
you got all the values , curves impulse response you want, but i guess something is missing then.

this is my message.
-listen and make some good sounding records with what you have
use the measurement tools to understand don't force the head to do stuff it cant handle.
-there are many variables. styli , blanks playback pickup....it gets tricky .
many of the best rated sounding records where done on absolutely non perfect Equipment "technically speaking" so there is hope

Caruso head has its limitations. this is due to its simple construction and its the price tag .
but it has incredible low end and for most of the stuff a really really great sound if you know how to handle.
with same preamp, amp you definitely will hear the difference of 7k-12k in price. sorry for that. there is limits.

seems amp has a big influence. the long suggested QSC GX3 looks great on schematics. there is even schematics already a plus.
after all these years. 14 years so far , we had to learn that the quality is variable from amp to amp.
not a surprise for the price tag. then way too much rail to rail voltage for most users. maybe modding to lower +/-45V will help already.
so i would strongly NOT recommend anymore.

there was a question also about what amp to use then:

after all these years i never got back a lot of information about great amps and what people are using.
i dont have the time to do the job for you . what i can tell is:

-PAS2002. german built amp sold at thomann a few years back.. on of the best.
-older 80/90 yamaha PA amps seem to work very well.
- flokason bameless amp IV. (a douglas self based design i use some derivate in ultracut since many years)
-LM3886 based amps just great (if you dont go too crazy with levels)

we all would love to get more amp suggestions. there must be many good ones out.
also the correction curves you use would help others to make it better.
so please share your experience and knowledge. there are at least 150 heads out and there should be some knowledge around

happy cutting.

[Dub Studio]

i would listen to the guy who designed and make/made the Caruso heads

Of course, I am not disputing anything Flo says, but I think its important to draw the distinction between using pink noise to help quantify and verify any adjustments you make and using pink noise to correct a cutting head to death, which is what (I believe) Flo was cautioning against, and rightly so.

A cutting head is, by its nature, hard to control. It has many moving parts, each with its own DOF, and no amount of correction can make all those parts behave exactly the way you want them to. The best we can hope for is to minimise unwanted artefacts as much as possible. Correction can help, but sometimes correction in one direction will create more problems in another. Hence why its a good idea to go easy, don't overdo it, and think of the cutting head as a complex system (not a simple algorithm).

Likewise I agree that using better equipment can render better results, but no system is immune to the above issues.

On the subject of amps, FWIW I use Bryston 3B-ST and they have powered my VRs adequately for many years without issue.

[leemichael]

On the subject of power amps, I have tried more than I can remember and so far the Bryston 4B ST and the 4B SST if you have the budget are the ones for me. The SST has the edge, goes a bit higher to 100khz and bit more power at 500 watts at 4 ohm where the ST has 400 watts.

The ST seems to sell for around £2k and the SST £2.5k used.

If anyone in the uk wants to upgrade from their QSC I could sell my 4B ST if interested.

Other amplifiers worth a mention not so expensive but pro audio touring amplifiers are the C Audio CR 404 / SR 606 / SR 707. Very capable, well built and meet most of the criteria however they’re getting a bit old now and probably need a service and some components renewed but they can be picked up for little money.

There’s others as well when I remember….

[flozki]

Bryston I can not recommend..the 2nd one on my negative list.
at least the ones mid 90-2000. They are super divas.. but maybe more recent ones are are a little more easy to handle

Not tried out since then..

[leemichael]

90-2000 is first generation, very old and not the same.
ST is 2000 onwards, SST around 10 years later, now on the SST2. All have major improvements on every update.