Proteins in Varnish Systems

[Michael_Molnar]

The issue about proteins in varnish systems is twofold. First, is it truly detected in old violins? Where is it found and in what concentration? Second, why was it used at all; that is, what do proteins do for a varnish system? By varnish system I mean the ground and wood preparation followed by layers of varnish and polishing compounds.

I have a couple of opinions on proteins. The first of which is an old theory that I come back to often. That idea is that the proteins were not intentional but a secondary component of an organic colorant or stain.  Many of you know that I have experimented a lot with annatto. These seeds are a major source of protein. Nevertheless, I no longer advocate annatto colorant. (I have an accepted publication in the VSA Journal on Annatto. I will give details later.) Anyhow, there are many colorants from seeds that contain protein. Itis possible that proteins can be carried along with a colorant. I will talk more about this some other day.

The other protein source is one that is intentional and applied for specific reasons. An example is a protein mixed with a colorant to stabilize and control it. That is what paint artist call a distemper which is similar to a tempera but without eggs. Or, a caseinate is applied to seal the wood.

… .

[MikeC]

why do you no longer advocate annatto?   

 

[JacksonMaberry]

I look forward to your paper, Mike. Thanks for sharing your work as always. 

I'm skeptical of the use of proteins deliberately, due to the cautious wording of Echard and Brandmair's findings. But we have all seen some very lovely instruments by colleagues which make use of it, and so one can't help but see it as a valid and viable approach. 

Looking forward to reading what others have to say. 

[Michael_Molnar]

The annatto paper, “Stabilizing Dichromatic Annatto Colorant”, is coming out of copyediting and will be published here:

J. Violin Soc. Am.: VSA Papers • 2022 • Vol. XXIX, No.1

I still use annatto but have concluded it is not historical for a Cremonese Instrument. Lupot (Paris 19th century) used annatto mixed with saffron. Evidently, some French makers still use it in a ground. Can anyone here cite makers and examples? Anyhow, once it is stabilized, annatto is incredibly beautiful. Its dichromacy produces golden (green-side) yellow, a range of orange shades, and dark reds. There is a tendency on my part to use too much which can make the dichromatic colors garish.

The point of my VSA paper is to explain something overlooked by luthiers who ignore art history. We need to understand why proteins could have ever been added intentionally. (Again, I am not claiming that proteins were intentionally added.) Proteinaceous mediums are common in art paint, and we see lots of luthier research suggesting their presence in a varnish system.
 

Proteins are more than a medium for conveyance. They can protect organic molecules from oxidation and UV radiation. I think they can also provide a barrier to destructive acidity in varnish. They can also levigate mineral colorants and provide an optical interface reducing the haze of light scattering. However, they can also be just a contaminate from the production of colorants. I noted their presence in seeds and nuts.

[Don Noon]

What about wetting?  Proteins usually are in water, which has a high surface tension and might not want to fully wet a rough surface (and at the microscopic level, wood is very rough).  It seems to me that good wetting is important for getting the best possible optical effects

[Andreas Preuss]

Mike, if you want to go the bottom of it, uhm, I mean to the ground, it is really not easy. 
 

i can’t answer if proteins were detected, that’s not my domain. For the group ‘old violins’ you certainly can detect proteins in the form of glue size on many old German instruments. However, I suppose you are aiming more on old Italian violins. 

I am more in the camp of violin makers who believes that plant stains were NOT used in the famous Cremonese ground. I base my opinion on the fact that the idea of wood treatments gets more and more a scientific fundament. These treatments were done IMO primarily for wood preservation and protection against pests but had the nice side effect of darkening the wood in a degree which made staining in any form not necessary. 
 

However, I think it is always good to re-examine the opposite view. (Instead of going into unnecessary and non constructive rants.)

Staining wood evenly and nicely is not easy and it is becoming more difficult on arched plates. So for this scenario I think protein layers can be very helpful for the coloring (almost regardless the used colorant). Very diluted Glue size or gelatin or casein can be used as pore blocker to ensure that the following color layer becomes evenly spread out on the surface. Additionally this makes a lot of sense because neither alcohol nor turpentine can dissolve such a layer and mess up the stain in the wood. 
 
(Experiments I had done on staining wood never rally satisfied me, but it’s possible that I never tried to refine any procedure.)
 
BUT…

… you can get also a very nice ‘stain’ on white wood with dark cooked varnish. 

[Andreas Preuss]

2 hours ago, Michael_Molnar said:

Lupot (Paris 19th century) used annatto mixed with saffron.

That’s quite interesting.

Actually, I have one question here. Can a protein layer protect the colors such as anatto and saffron from fading? Both are not so lightfast. 

[David Beard]

From a generic historical perspective, almost every surface preparation in the arts and particularly the Italian arts involved 'sizing' the surface at an early stage of work.

'Sizing' was a near universal preparation stage to every surface.   Reasons given for sizing where to unify and improve the integrity of a surface, and to give a uniform 'tooth' to the surface.

'Sizes' were based on a thinned application of some binder.   Many things served as 'sizing' binders in special cases, from garlic to oil to rosin.  But, the general purpose dominant binder was protein, either egg or animal skin.

********

Given such context, the absence of protein as a sizing binder would be more surprising and remarkable than its presence.

[John Harte]

3 hours ago, Andreas Preuss said:

i can’t answer if proteins were detected, that’s not my domain. For the group ‘old violins’ you certainly can detect proteins in the form of glue size on many old German instruments. However, I suppose you are aiming more on old Italian violins.

It seems that proteins have been detected by various researchers in almost all varnish systems that have been studied including old Cremonese.  The big question is the nature or form of the protein presence.  Like the German instruments you mention, various old French and Italian (including some Cremonese) have been identified as having a glue size present within their varnish systems.  There are also instruments by various makers where the situation seems less obvious where no discrete protein based layer has been conclusively identified.

[John Harte]

17 minutes ago, David Beard said:

From a generic historical perspective, almost every surface preparation in the arts and particularly the Italian arts involved 'sizing' the surface at an early stage of work.

'Sizing' was a near universal preparation stage to every surface.   Reasons given for sizing where to unify and improve the integrity of a surface, and to give a uniform 'tooth' to the surface.

'Sizes' were based on a thinned application of some binder.   Many things served as 'sizing' binders in special cases, from garlic to oil to rosin.  But, the general purpose dominant binder was protein, either egg or animal skin.

********

Given such context, the absence of protein as a sizing binder would be more surprising and remarkable than its presence.

I agree.  And this is the problem that the research community considering violin varnish systems face.  They expect this to be the case, often limiting their focus.  Confirmation bias has been an ongoing problem and remains so.

[David Beard]

Yes.   And, the historical context offers few or no examples that skip the sizing role when a process is described.

But, there are quite a few where something other than protein is used to size.

So, in terms of historical context, using balsaam or oil or rosin as sizing is more reasonable than omitting the sizing role entirely.

 

[John Harte]

17 minutes ago, David Beard said:

Yes.   And, the historical context offers few or no examples that skip the sizing role when a process is described.

But, there are quite a few where something other than protein is used to size.

So, in terms of historical context, using balsaam or oil or rosin as sizing is more reasonable than omitting the sizing role entirely.

 

By the very nature of varnish systems, there has to be some form of initial application for which the term sizing effectively applies.  The reasons behind whatever is chosen may vary.  I don't think that many would argue with this.  On the other hand, the question posed in this thread is specific to proteins in varnish systems and, I guess, the nature of these.

[John Harte]

6 hours ago, Don Noon said:

What about wetting?  Proteins usually are in water, which has a high surface tension and might not want to fully wet a rough surface (and at the microscopic level, wood is very rough).  It seems to me that good wetting is important for getting the best possible optical effects

I agree.  This is one reason why I am not surprised that researchers have been unable to find evidence of an observable discrete protein based layer on the wood surface and/or within the upper wood cell structure in certain instruments in spite of there being a detected protein presence within their varnish system. 

[David Beard]

39 minutes ago, John Harte said:

I agree.  This is one reason why I am not surprised that researchers have been unable to find evidence of an observable discrete protein based layer in certain varnish systems in spite of there being a detected protein presence. 

I agree with this also.  Better wetting, for optical reasons, is exactly why I would expect some or perhaps even most of the old makers to move to using a balsaam/rosin/or oil size that would wet better initial and even continue to progress wetting for years.

However, I would expect them to still conceive this step as 'sizing'.  This means applying the 'size' thinned, then letting it dry.  Applying warm or hot is also a possibility.   And, as with all sizing, the aim would be to 'well touch' and unify the structure, but to 'close' the structure.   That is to wet, but not clog.

This is I believe the benefit of understand this step of work as 'sizing', to wet but not clog.

 

Also, I would not expect consistent details to the processes between makers, or even across different examples from one maker.

The culture was to consistently size.  And sizing consistently meant to bind, wet, and not clog.   But, to size with protein or with something more transparent, or with a thin emulsion, etc.  These choices of process detail might vary.

 

Perhaps this is part of way test results seem to vary?  Perhaps makers sometimes used proteins, but sometimes didn't.

[John Harte]

15 minutes ago, David Beard said:

Perhaps this is part of way test results seem to vary?  Perhaps makers sometimes used proteins, but sometimes didn't.

Quite possible.  However there still exists the issue of the exact nature of protein presence in those instruments that don't seem to posses an observable discrete protein based size (e.g., casein, hide glue etc.,) and yet exhibit a detected protein presence.  This seems to be the case, for example, with almost all Strad varnish micro-sample material.

[Don Noon]

22 minutes ago, David Beard said:

And, as with all sizing, the aim would be to 'well touch' and unify the structure, but to 'close' the structure.   That is to wet, but not clog.

I think the phenomenon of "glue ghosts" (i.e. low contrast flames or a "flat," pale appearance) is the result of a coating that seals the wood with poor wetting or penetration.  In my experiments, casein/ammonia will do that too, if applied thickly enough.  If applied very thinly, it doesn't seem to seal the wood from subsequent coatings, and can (at best) be optically insignificant.

The main attraction of protein coatings (at least for me) is that they can be very low damping, so I use it on the inside, but not on the outside where optics is important.  A resin-solvent ground/sealer can also be low damping, and get good optics.

[John Harte]

5 minutes ago, Don Noon said:

I think the phenomenon of "glue ghosts" (i.e. low contrast flames or a "flat," pale appearance) is the result of a coating that seals the wood with poor wetting or penetration.  In my experiments, casein/ammonia will do that too, if applied thickly enough.  If applied very thinly, it doesn't seem to seal the wood from subsequent coatings, and can (at best) be optically insignificant.

Being incredibly picky here....  In my experience, even very thin low concentration applications have a tendency to veil or interrupt the ability to see the full detail of wood structure.  Even emulsions do this.  This can be incredibly subtle but under certain viewing conditions it always seems to be there... 

[jezzupe]

7 hours ago, Don Noon said:

I think the phenomenon of "glue ghosts" (i.e. low contrast flames or a "flat," pale appearance) is the result of a coating that seals the wood with poor wetting or penetration.  In my experiments, casein/ammonia will do that too, if applied thickly enough.  If applied very thinly, it doesn't seem to seal the wood from subsequent coatings, and can (at best) be optically insignificant.

The main attraction of protein coatings (at least for me) is that they can be very low damping, so I use it on the inside, but not on the outside where optics is important.  A resin-solvent ground/sealer can also be low damping, and get good optics.

My take on this is that many may opt to do a glue wash to eliminate glue ghosts, and or contrasting sealer colors.

 I think the real concern or issue is proportion and that any glue/protein wash needs to be very thin so as to allow  solvent finish loading/wetting in the grain on top of the very thin protein layer.

When an instrument is done, if for some reason I know I may be dealing with some "invisible slop" from some "gluing incident" I will not hesitate to whip up a VERY light glue water mix that I will work into the grain in order to resolvent the invisible glue ghosts and in essence apply a very light "water based COLOR sealer...ie. pale glue cast UNDER the first varnish layers. This will eliminate any un sightly glue ghosts and eliminate the worry of applying varnish only to have them pop up.

Glue absorption spots can be dealt with in many ways

1. don't get them in the first place, this can be easier said than done

2. if you do get them, you must eliminate ALL of them if you are going to go straight to a solvent based ground, elsewise white spots

3.This can lead to some "guessing" as far as "did I get them all" if not ,white spots, and in certain circumstances one might not be able to surface prep deep enough to remove them all, and then if you do get them, AFTER  a solvent coat, you are then forced to either live with them, or to try to remove them with scraping.

4, So many may opt to just eliminate all the guessing {depending on their neatness} and apply a glue wash so as to not worry "if they got them all"

So to me it may be a step done by many now and in the past in order to eliminate dealing with glue ghosts entirely by applying a light glue ghost to the entire instrument, be it a glue wash/ casein or glair , the effect of coloration will be similar.

So to me any "glue/casien or glair based coating must be "thinish in order to winish" lest we get checking and milkyness down the road

But I do like it when I KNOW I've done a very clean job and that I won't be seeing ghosts and don't need to do a wash...BUT I certainly can see the logic of using them as sealers to eliminate the guess work and headaches of trying to fix the ghosts.

some of the worst ones I have got were not from application slop but from having dry glue on my hands and forgetting to wash my hands, and then handling the instrument in summer weather, and then only after holding it in a certain way for a long period of time did I notice my hands were tacky from body heat contact, this was a mahogany guitar, but I thought, it won't be an issue, but it was, so I had many spots to address.

It's also another reason I like sugar ground because it will resolvent any tiny glue ghosts and then the sugar wash will wipe them off/in ,make them not noticeable and the grain won't pop when dry because the sugar slows the water absorption rate.

 

[Michael Szyper]

IMO It is unlikely that cremonese used annatto as a wood stain. First, it is a positive stain (not hiding Winter growth in maple). Second, it has a orange uv-fluorescence. Cremonese wood usually has a fluorescence extinction in the upper layers.

 Nevertheless, I second that it looks beautiful. 

[Michael_Molnar]

1 hour ago, Michael Szyper said:

IMO It is unlikely that cremonese used annatto as a wood stain. First, it is a positive stain (not hiding Winter growth in maple). Second, it has a orange uv-fluorescence. Cremonese wood usually has a fluorescence extinction in the upper layers.

 Nevertheless, I second that it looks beautiful. 

Could you explain a little more about “positive stain”? 

[scordatura]

1 hour ago, Michael Szyper said:

IMO It is unlikely that cremonese used annatto as a wood stain. First, it is a positive stain (not hiding Winter growth in maple). Second, it has a orange uv-fluorescence. Cremonese wood usually has a fluorescence extinction in the upper layers.

 Nevertheless, I second that it looks beautiful. 

UV is a great way to check if you are going in the right direction for Cremonese varnish adventures. The only variation between old and new is oxidation.

[Michael Szyper]

51 minutes ago, Michael_Molnar said:

Could you explain a little more about “positive stain”? 

I think this is a German translated term of wood stains. Brandmair mentions that as well. Negative stains are most common - colorants which color summer growth a lot more than the Winter growth - resulting in a negative looking picture: Winter rings being brighter than the summer growth. Positive stains do not change the typical appearance of wood with dark winter rings. I am quite convinced that in the case of annatto it is the lye you need to react bixin -> norbixin for making it water soluble. Lyes are quite strong positive stains. On cremonese fiddles I See negative stain patterns on maple and positive staining patterns on spruce. 

[Michael Darnton]

2 hours ago, Michael Szyper said:

IMO It is unlikely that cremonese used annatto as a wood stain. First, it is a positive stain (not hiding Winter growth in maple). Second, it has a orange uv-fluorescence. Cremonese wood usually has a fluorescence extinction in the upper layers.

 Nevertheless, I second that it looks beautiful. 

Annatto is beautiful stuff, but highly fugitive. I once made a lake pigment with it in hopes of protecting it, everyone agreed that the color was gorgeous, and after one night in the light box it was entirely gone.

[Don Noon]

6 hours ago, jezzupe said:

My take on this is that many may opt to do a glue wash to eliminate glue ghosts, and or contrasting sealer colors.

Does anyone else think this is a lame way to avoid glue ghosts... by washing out the natural contrast of the entire violin?

1 hour ago, scordatura said:

The only variation between old and new is oxidation.

To be picky, oxidation isn't the only thing going on, color-wise.  Torrefied wood is processed in an oxygen-free atmosphere, so it is not oxidized.  And UV does things too (although I don't know the chemistry of this, or if oxygen is involved).

[Michael Szyper]

Just now, Michael Darnton said:

Annatto is beautiful stuff, but highly fugitive. I once made a lake pigment with it in hopes of protecting it, everyone agreed that the color was gorgeous, and after one night in the light box it was entirely gone.

After coating the stained wood with a clear varnish coat it seems to be pretty stable. Generally I could not observe that lakes are more lightfast than their pure colorants. But lot of respected people would disagree.