I'm Interested in a Microscope

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

sidepart

Well-Known Member
Joined
Jun 1, 2012
Messages
294
Reaction score
33
Location
Saint Paul
I've been trying to justify purchasing a microscope for a long time, and this might just be the justification I'm looking for. I just need a recommendation on the specs (zoom, style, etc).

Here's what I'm looking at right now: http://www.amazon.com/dp/B000NOU54O/?tag=skimlinks_replacement-20

Here's the story:

I racked one of my brews last night and the sample I tasted hit me as a bit sour. Totally unexpected for the recipe and it had me concerned. I was out of the house last week (honeymoon) and figure that the basement temperature just got too hot. WLP007 at higher temperatures does produce a more sour flavor. Still...maybe there's some Lactobacillus afoot in my brew, and I'd love to see it. The scientific interest is too much for me to pass up here.
 
Last edited by a moderator:
I think 400x is the minimum for observing yeast. Bacteria are smaller. I'm not sure how much. I think 400x would still be useful but higher magnifications sure would be helpful. I think I've heard binocular styles are better. I've been thinking about a microscope also. I keep wondering if surplus lab grade units are available anywhere.
 
I spend a lot of time on a microscope- The glass is the most important thing-great glass= great imagine, binocular styles are better but unless you need to spend a great deal of time looking monocular works fine (hold your hand over your other eye to reduce fatigue). I do think it is possible to see the various yeasties and bacteria at lower than 100X, check here
http://sciencebrewer.com/2011/07/20/wild-yeast-project-lambic-plating-update-bacteria/

I just got a hemocytometer and really was cool and plan to ranch and freeze soon. You might wanna check ebay for something used because who knows how good those objectives/ocular are from that amazon source. Now what to do with the 3 old microscope bodies I have....

Good luck,
Dan B.
 
A microscope is one of those things which requires that it be well made in order for it to really be useful. As a rule of thumb I'd say anything costing less than about $1500 new would turn out to be a piece of junk - a splendid example of the old saw that the bitterness of poor quality remains long after the sweetness of low price is forgotton.

Obviously, you don't want to spend $1500 so this is where jeepinjeepin's recommendation comes in. See if you can find a good used microscope. There are lots of ways to screw up a microscope, of course, but optics don't wear out by passing a lot of light through them so if the thing hasn't been dropped off the bench or had corrosive acid poured over it it should perform perfectly well.

As for magnification: for yeast counting 40 - 60x is fine. For being able to distinguish various yeast strains and identify bacteria 400x should suffice. If higher magnification is desired then one gets into oil immersion which is a real pain. Yes, you can buy microscopes advertising 1000X but without oil you get what is called 'empty magnification'. A yeast cell looks bigger but it's a big blur.

To really see into yeast cells phase contrast is needed (this involves a special condenser and set of lenses but can often be retrofitted to a garden variety bright field scope).

I guess I'm saying money spent on a First Lab Duo Scope would be money wasted.
 
ajdelange said:
As for magnification: for yeast counting 40 - 60x is fine. For being able to distinguish various yeast strains and identify bacteria 400x should suffice.

Thanks. I wasn't sure about it. I knew I had heard the 400x number. Glad to hear that is the upper end needed.
 
I just found a half dozen or so on Craigslist near me at $300-500. One was even a Nikon SC with 1000x oil immersion lens. I do not have a disposable $500 though.
 
Just checked ebay..... plenty of AO Spencer, Fisher, Swift, Leitz and Wesco (I never used Wesco) for $100 or less.
 
I respect A.J's insight a lot, but I don't agree that a brewer's microscope has to cost $1500.

What we are using the microscpope would be:
- cell counting
- looking at yeast in general
- identifying haze
- possibly identifying contaminants

The latter is tricky since once there are enough bacteria so you see them in a sample your beer should have a significant off flavor already.

A few years back I bought a up to 1000x (100X oil immersion objective and 10x eyepiece) monocular student microscope new for $160 from Cynmar and I'm happy with it.

at 1000x yeast look like this:

yeast_1000x_and_zoom.JPG


Sure, the pic may have been better with a $1500 scope, but that doesn't detract from its usefulness.

These days I generally count yeast from cell-phone photos like this. The image is blurred on the edges b/c I just held the phone camera to the eyepiece. This is not how I see the cells when I look through the microscope.

I do cell counts with ImageJ:

ImageJ_cell_counting.jpg


Kai
 
If you are happy with the $150 scope then that's all there is to it. But I think your money might well have been better spent on finding an old instrument, at or near that price, of higher quality. The 1000X picture in #9 is a perfect example of 'empty magnification' for which hobbyist telescopes and microscopes are infamous. The cells are large but it is impossible to see any of the details of their interiors. It therefore doesn't matter how big they are. They might as well be little dots. The fact of their being bigger gains you no additional visual information.

Getting real (non empty) magnification isn't easy. The resolution of an optical instrument depends on its numerical aperture. This is the index of refraction times the sin of the light acceptance angle which in turn depends on the focal length and physical aperture (lens diameter) i.e. on the f/ number of the lens. Oil gets you a boost in NA through higher index of refraction but you still have to solve the problem of low f/ number with decent MTF, low distortion etc. This is where the cheapie scopes fall flat. Then the marketing department zooms in and says they have to offer 1000x with immersion because the competitor does. It's like the pixels in a digital camera. Everyone knows more is better but never thinks that if the supplied lens doesn't support the pixel pitch the extra pixels are useless.

With telescopes there is a simple rule of thumb. Any magnification above x times the aperture (I don't remember what x is) is a waste. It ought to be easy to map that over into y times the NA of a microscope objective. But this assumes that the MTF of the lenses is largely limited by diffraction - not poor grinding or other manufacturing shortcomings.

If the only application is yeast counting then clearly magnification way less than 1000X is fine and a student microscope should suffice. It is when one wishes to detect bacteria and/or look at yeast cell morphology (with a little training you can recognize different strains) that real resolution is required.

There is a reason that the lens for your HD video camera costs $1000 but a real cine lens costs 20 times that. Optics is, despite the tremendous advantages lent by computer ray tracing, still very much an art.

The imageJ software looks pretty cool. I've downloaded the users guide and will probably download the software.
 
I bought a used Olympus CH off ebay, ~$500. For anyone interested in a used scope, the CH/CH2 or BH/BH2 models are great, sturdy scopes. Olympus makes a whole array of objectives of different quality for these scopes. The objectives I got seem to be suitable for some cell morphology study (vacuoles, wall struction, bud scars, etc are visible under 400x. Mitocondria, nucleous, etc are not so much; prolly need 1000x and oil immersion for that, which unfortunately I don't have). Fluorescence and especially phase contrast options are remarkable but also add significant cost to the scope.

When I was looking, it seemed that the safest bets were Olympus, Nikon, Zeiss, & Leica. Unless you have a lot of experience with this sort of thing, you are gambling when you buy a off-brand scope, even new.
 
....Olympus, Nikon, Zeiss, & Leica.

Too many Japanese and Germans. I was going to point out that we have our own Germans right here in the USA - Bausch and Lomb and that got me wondering as to whether they still make microscopes (my urologist has a Galen - the same model I have) but they don't. They sold all that off and now make contact lenses. However, when I went and looked at the link in #8 I found out to whom they apparently sold the microcscope designs - that Olympus is a dead ringer for my Galen.

So if you want a good, solid German microscope designed in America and made in Japan you can look for an Olympus.
 
agreed, the 1000x magnificantion is lacking for the scope that I have. But then again, I didn't buy it for that magnification and the 400x looks just fine. It's not that I didn't want a nicer quality one, I just didn't want to spend too much and didn't feel like watching Ebay or Craigslist.

An upgrade that would really help me work more efficiently would be an attached camera. I have taken brief looks on the internet and even 5MP microscope cameras are pretty expensive. At least for my budget. I also don't know what the picture quality will be. An attachment for my DLSR would also be an option, but not a solution that I could keep on the microscope and plugged into the computer.

I also found that counting cells adds a lot of time to the brewday. Because of that I have started establishing a correlation between yeast sediment weight and cell count. This way I can pitch by weight and don't worry about counting cells all the time. Or I pitch by weight (under the assumption of a particular sediment density), take cell count pictures and count later to verify the assumption.

ImageJ is nice. Unfortunately the pictures I'm taking are not lit evenly enough so I can thresholding and the automated cell count feature. But that will have trouble with trub and budding cells anyway. Before ImageJ I would actually use Paint to place dots on cells I counted with a tally counter. Before that I counted with a tally counter while looking through the scope.

Kai
 
agreed, the 1000x magnificantion is lacking for the scope that I have. But then again, I didn't buy it for that magnification and the 400x looks just fine. It's not that I didn't want a nicer quality one, I just didn't want to spend too much and didn't feel like watching Ebay or Craigslist.

An upgrade that would really help me work more efficiently would be an attached camera. I have taken brief looks on the internet and even 5MP microscope cameras are pretty expensive. At least for my budget. I also don't know what the picture quality will be. An attachment for my DLSR would also be an option, but not a solution that I could keep on the microscope and plugged into the computer.

I also found that counting cells adds a lot of time to the brewday. Because of that I have started establishing a correlation between yeast sediment weight and cell count. This way I can pitch by weight and don't worry about counting cells all the time. Or I pitch by weight (under the assumption of a particular sediment density), take cell count pictures and count later to verify the assumption.

ImageJ is nice. Unfortunately the pictures I'm taking are not lit evenly enough so I can thresholding and the automated cell count feature. But that will have trouble with trub and budding cells anyway. Before ImageJ I would actually use Paint to place dots on cells I counted with a tally counter. Before that I counted with a tally counter while looking through the scope.

Kai

I have a very nice solution now, but a while ago I took apart a cheap webcam and used it as a camera. You must remove the webcam lens, which might be a chore, but it can be done. After the webcam lens is removed, remove the eyepiece lens of your microscope (typically 10x). Mount the webcam into the open eyepiece hole. It works great. Cheers.
 
I have a very nice solution now, but a while ago I took apart a cheap webcam and used it as a camera. You must remove the webcam lens, which might be a chore, but it can be done. After the webcam lens is removed, remove the eyepiece lens of your microscope (typically 10x). Mount the webcam into the open eyepiece hole. It works great. Cheers.

Thanks. I have to give that a try.
Do you have sample images?

Kai
 
ajdelange said:
Too many Japanese and Germans. I was going to point out that we have our own Germans right here in the USA - Bausch and Lomb and that got me wondering as to whether they still make microscopes (my urologist has a Galen - the same model I have) but they don't. They sold all that off and now make contact lenses. However, when I went and looked at the link in #8 I found out to whom they apparently sold the microcscope designs - that Olympus is a dead ringer for my Galen.

So if you want a good, solid German microscope designed in America and made in Japan you can look for an Olympus.

I'm all for Germans and Japanese when it comes to optics. I wish more of my Nikon lenses were the upper level stuff that are actually made in Japan instead of elsewhere.
 
ImageJ is nice. Unfortunately the pictures I'm taking are not lit evenly enough so I can thresholding and the automated cell count feature. But that will have trouble with trub and budding cells anyway. Before ImageJ I would actually use Paint to place dots on cells I counted with a tally counter. Before that I counted with a tally counter while looking through the scope.

There are lots of things you could try to fix this. A simple gaussian blur followed by a top-hat filter (think this gets implemented as the 'Subtract Background' function in ImageJ) will probably get you close enough so you can threshold and do simple cell counts as long as you don't need to be extremely precise. You can also try the 'Analyze Particles' function to deal with trub and other 'noncircular' things. There's an option there to only count thresholded regions that meet a certain degree of circularity. You could also play around with watersheds to separate cells that are touching.

Beyond that, automated cell counting has been heavily looked at by the image processing world. Plenty of more sophisticated algorithms exist but finding ImageJ ports for these might be tricky.
 
Back
Top